What on earth are tech companies building all these data centers for? It turns out, they’re not just to help us search for recipes or plan our vacations. Simple chatbot queries are now an outdated way of thinking about AI. The real deal is in ‘agents’—large language model-based systems designed to make autonomous decisions. These agents can run for hours, re-prompting themselves to build different web pages, menus, and datasets, all in the name of solving complex tasks. But the cost? A lot of energy: probably tens of millions of dollars' worth.

Private AI companies have historically been choosy about what to disclose when it comes to environmental metrics. Many CEOs point to single queries made by individuals as a measure of resource use, but introducing agents, which are much more energy-intensive, makes these calculations a lot more complex. There’s a major dearth of information around the energy use of agents, whose tasks can range from simple jobs to a full day of autonomous coding involving a team of parallel ‘helper’ agents.

Some AI enthusiasts are trying to do the math themselves. Climate scientist Zeke Hausfather calculated that his average daily Claude session may consume more than the energy needed to power two refrigerators. But this represents a net new source of emissions, at a time when global temperatures are skyrocketing and our emissions reduction goals are increasingly off track. And it’s a lot bigger than the fraction-of-an-almond-sized numbers Sam Altman is throwing around.

As AI companies rush to build data centers to accommodate these power-hungry agents, the future looks like a world with hundreds or even thousands of AI agents working in the background. This could be the way we live in the future, but what does that mean for our energy consumption and the environment? The rush to build data centers is not just about technology—it’s about power.